optistruct optimization

Hüsnü Yalınkılıc

hi dear moderators,

I conducted a gauge optimization study by the optistruct, but it completed by just a iterations and it had said the design was infeasible. how is possible?

what is problem or what should I do ?

thank you for quick reply PLS help

Gildas GUILLY_21156

Hello,
Your optimization did just 2 iterations but no changes were done to the design, in particular I guess because you have some constraints that just cannot be verified.
If we look at the worst one : you have constrained an acceleration to 0.4 while its current value is 2.61, meening you hope your optimziation will be able to divide the acceleration by a factor >5.
Are you sure the variables you have defined in your model (adding a maximum of 20% thickness on 2 properties) could be enough to solve this constraint ?

To sum up, I think this is a problem in your otpimziation definition.
can you provide more details about what you are trying to acheive ?

Gildas

Hüsnü Yalınkılıc

hi Guilly, 

thank you for quick answer

Your optimization did just 2 iterations but no changes were done to the design, in particular I guess because you have some constraints that just cannot be verified.

I am wondering about it too, why it finished at the second iteration if the optimization continued, maybe this constrains would be verified.

Are you sure the variables you have defined in your model (adding a maximum of 20% thickness on 2 properties) could be enough to solve this constraint ?

if the thickness values decrease my constrained will be verified I m sure but the thickness increased  

Gildas GUILLY_21156

Your thicknesses do not increase, they stay at 1e-3 on both iterations (see design property tables in your out file)
Are you sure decreasing your thickness would solve ALL violated constraints ?
In particular you have constraints on static displacement dans modal frequency.
reading at just your out file, my understanding is that you have defined an impossible optimization problem.

Gildas

Hüsnü Yalınkılıc

Are you sure decreasing your thickness would solve ALL violated constraints ?

yes I'm sure, look me

if thickness decrease modal frequencies decrease and the deflections in the structure because decreasing in the stiffness, absolutely

Your thicknesses do not increase, they stay at 1e-3 on both iterations (see design property tables in your out file)

in the first iteration why thickness values have not changed. I think this situation is absurd.

how the algorithm decides on thickness values? @Gildas GUILLY 

thank you

 

 

 

Gildas GUILLY_21156

As I said, I think the algorythm do not identify variations of the thicknesses that allow to solve the different contraints.
Can you share your fem file ?
If not (confidential) I would suggest you contact Altair support where you will be able to share your model file for investigation.

Gildas

Hüsnü Yalınkılıc

As I said, I think the algorythm do not identify variations of the thicknesses that allow to solve the different contraints.

it is running now, and thickness values decreased in the second iteration

if you want the model, it was attached